CN111460732A - Construction method of nonlinear model of planar motor - Google Patents

Construction method of nonlinear model of planar motor Download PDF

Info

Publication number
CN111460732A
CN111460732A CN202010243219.XA CN202010243219A CN111460732A CN 111460732 A CN111460732 A CN 111460732A CN 202010243219 A CN202010243219 A CN 202010243219A CN 111460732 A CN111460732 A CN 111460732A
Authority
CN
China
Prior art keywords
model
planar motor
nonlinear
neural network
state information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202010243219.XA
Other languages
Chinese (zh)
Other versions
CN111460732B (en
Inventor
黄苏丹
胡智勇
曹广忠
陈龙
孙俊缔
敬刚
刘岩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen University
Original Assignee
Shenzhen University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shenzhen University filed Critical Shenzhen University
Priority to CN202010243219.XA priority Critical patent/CN111460732B/en
Publication of CN111460732A publication Critical patent/CN111460732A/en
Application granted granted Critical
Publication of CN111460732B publication Critical patent/CN111460732B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Control Of Electric Motors In General (AREA)

Abstract

The invention discloses a construction method of a nonlinear model of a planar motor, which is used for establishing a nonlinear dynamic model of the planar motor and establishing a neural network model based on the nonlinear dynamic model of the planar motor; training the neural network model based on a preset training sample set; and taking the model parameters of the trained neural network model as the model parameters of the nonlinear dynamic model of the planar motor to obtain the nonlinear dynamic model of the planar motor. The plane motor model is a nonlinear dynamic model, reflects the nonlinear dynamic characteristics of the plane motor and has high model precision; the model parameters of the nonlinear model of the planar motor are solved through the neural network model, so that the model reliability is improved, and the method can be used for a controller of the planar motor to improve the accuracy of the position control of the planar motor.

Description

Construction method of nonlinear model of planar motor
Technical Field
The invention relates to the technical field of planar motors, in particular to a construction method of a nonlinear model of a planar motor.
Background
The planar motor has the advantages of simple structure, convenience in installation, good heat dissipation, high precision, high speed, low cost, high reliability and the like, and has great application prospect in the field of precision manufacturing. At present, high-precision position control is a key point of attention in the field of planar motors, the accuracy of a mathematical model of the planar motor seriously influences the high-precision operation of the mathematical model, and in addition, the high nonlinear characteristic of the planar motor greatly hinders the precise modeling of the mathematical model.
Disclosure of Invention
The invention aims to solve the technical problem of providing a method for constructing a nonlinear model of a planar motor aiming at the defects of the prior art.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a method of constructing a nonlinear model of a planar motor, the method comprising:
establishing a nonlinear dynamic model of a planar motor, and establishing a neural network model based on the nonlinear dynamic model of the planar motor, wherein model parameters of the neural network model are model parameters of the nonlinear dynamic model of the planar motor;
training the neural network model based on a preset training sample set;
and taking the model parameters of the trained neural network model as the model parameters of the nonlinear dynamic model of the planar motor to obtain the nonlinear dynamic model of the planar motor.
The method for constructing the nonlinear model of the planar motor comprises the steps that the preset training sample set comprises a plurality of training samples, each training sample comprises training data and real state information corresponding to the training data, and the training data comprises state information and control quantity.
The method for constructing the nonlinear model of the planar motor is characterized in that the real state information is state information of the training data at the next moment, wherein the state information comprises speed information and position information.
The method for constructing the nonlinear model of the planar motor, wherein training the neural network model based on a preset training sample set specifically comprises:
inputting the training data into the neural network model, and extracting the prediction state information of the training data through the neural network model;
determining a loss function corresponding to the training data based on the predicted state information and the real state information corresponding to the training data;
training the neural network model based on the loss function.
The construction method of the nonlinear model of the planar motor is characterized in that the neural network model comprises a mapping unit and a linear transformation unit; inputting the training data into the neural network model, and extracting the predicted state information of the training data through the neural network model specifically comprises:
inputting the state information into the mapping unit, and outputting a coupling parameter through the mapping unit;
and inputting the coupling parameters and the control quantity into the linear transformation unit, and outputting the prediction state information of the training data through the linear transformation unit.
The construction method of the nonlinear model of the planar motor comprises the following steps of:
Figure BDA0002433244320000021
wherein J is a loss function, xreal_l(k) For true state information, xl(k) To predict the state information, k denotes the kth time.
The establishment of the nonlinear dynamic model of the planar motor comprises the following steps:
Figure BDA0002433244320000022
wherein ,
Figure BDA0002433244320000023
xl1(k) position information at time k on the l-axis, xl2(k) Speed information at time k on the l-axis, yl(k) Position information output for the model at time k, Cl＝[1 0]，Gl and HlIs a coefficient matrix of the l-axis, Fl[xl(k)]A non-linear state vector of the l-axis, ul(k) In order to control the amount of the liquid,
Figure BDA0002433244320000024
xl1(k +1) is position information of the time k +1 on the l-axis, xl2(k +1) is velocity information at the time of the l-axis k + 1.
A control method of a planar motor is applied to the planar motor nonlinear model constructed by the construction method of the planar motor nonlinear model, and the method comprises the following steps:
determining expected position information of the planar motor at the next moment based on the nonlinear model of the planar motor;
determining a control amount of the planar motor based on the expected position information and actual position information of the planar motor;
and applying the control quantity to the planar motor to control the planar motor.
A computer readable storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement the steps in the method for constructing a nonlinear model of a planar motor as described in any one of the above.
A terminal device, comprising: a processor, a memory, and a communication bus; the memory has stored thereon a computer readable program executable by the processor;
the communication bus realizes connection communication between the processor and the memory;
the processor, when executing the computer readable program, implements the steps in the method for constructing a nonlinear model of a planar motor as described in any one of the above.
Has the advantages that: compared with the prior art, the invention provides a construction method of a nonlinear model of a planar motor, which is used for establishing the nonlinear dynamic model of the planar motor and establishing a neural network model based on the nonlinear dynamic model of the planar motor; training the neural network model based on a preset training sample set; and taking the model parameters of the trained neural network model as the model parameters of the nonlinear dynamic model of the planar motor to obtain the nonlinear dynamic model of the planar motor. The plane motor model is a nonlinear dynamic model, reflects the nonlinear dynamic characteristics of the plane motor and has high model precision; the model parameters of the nonlinear model of the planar motor are solved through the neural network model, so that the model reliability is improved, and the method can be used for a controller of the planar motor to improve the accuracy of the position control of the planar motor.
Drawings
Fig. 1 is a flowchart of a method for constructing a nonlinear model of a planar motor according to the present invention.
Fig. 2 is a schematic diagram of a planar motor control system in the method for constructing a nonlinear model of a planar motor according to the present invention.
Fig. 3 is a schematic diagram of a training process of a neural network model in the method for constructing a nonlinear model of a planar motor according to the present invention.
Fig. 4 is a structural schematic diagram of an embodiment of a neural network model in the method for constructing a nonlinear model of a planar motor according to the present invention.
Fig. 5 is a schematic structural diagram of another embodiment of a neural network model in the method for constructing a nonlinear model of a planar motor according to the present invention.
Fig. 6 is a schematic structural diagram of a terminal device provided in the present invention.
Detailed Description
The invention provides a method for constructing a nonlinear model of a planar motor, which is further described in detail below by referring to the attached drawings and embodiments in order to make the purposes, technical schemes and effects of the invention clearer and clearer. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.
It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
The embodiment provides a method for constructing a nonlinear model of a planar motor, which can be applied to electronic equipment with front-end camera shooting or rear-end camera shooting functions, and the electronic equipment can be realized in various forms. Such as a cell phone, a tablet computer, a palm top computer, a Personal Digital Assistant (PDA), etc. In addition, the functions realized by the method can be realized by calling the program code by a processor in the electronic equipment, and the program code can be saved in a computer storage medium.
As shown in fig. 1, the present implementation provides a method for constructing a nonlinear model of a planar motor, which may include the following steps:
s10, establishing a plane motor nonlinear dynamic model, and establishing a neural network model based on the plane motor nonlinear dynamic model, wherein model parameters of the neural network model are model parameters of the plane motor nonlinear dynamic model.
Specifically, the nonlinear dynamic model of the planar motor is a model established based on the control quantity and the state information, and the nonlinear dynamic model of the planar motor comprises a state equation and an output equation. The state equation is used for predicting state information of the planar motor at the next moment, and the output equation is used for outputting position information of the current moment. The control quantity is an output quantity of a controller in a planar motor control system (for example, the planar motor control system shown in fig. 2), and the state information includes speed information and position information.
The nonlinear dynamic model of the planar motor established based on the control quantity and the state information can be as follows:
Figure BDA0002433244320000051
wherein ,
Figure BDA0002433244320000052
xl1(k) position information at time k on the l-axis, xl2(k) Speed information at time k on the l-axis, yl(k) Position information output for the model at time k, Cl＝[1 0]，Gl and HlIs a coefficient matrix of the l-axis, Fl[xl(k)]A non-linear state vector of the l-axis, ul(k) In order to control the amount of the liquid,
Figure BDA0002433244320000053
xl1(k +1) is position information of the time k +1 on the l-axis, xl2(k +1) is velocity information at the time of the l-axis k + 1.
Further, in one implementation of this embodiment, the non-linear state vector F of the l-axisl[xl(k)]The nonlinear characteristic is used for representing the nonlinear characteristic of the nonlinear dynamic model of the planar motor, the nonlinear dynamic characteristic of the planar motor is reflected through the nonlinear characteristic, and the model precision is improved. Wherein the non-linear state vector F of the l-axisl[xl(k)]May be:
Figure BDA0002433244320000054
wherein ,Al1，Al2And alGain coefficients, D, all on the l-axisl[xl1(k)] and Dl[xl12(k)]A coupling function of the l-axis for representing the coupling characteristics of the nonlinear dynamical model of the planar motor, Dl[xl(k)] and Dl[xl12(k)]Expressed as:
Dl1[xl(k)]＝ωl1·xl(k)＝ωl11xX1+ωl12xX2+ωl13xY1+ωl14xY2
Dl2[xl(k)]＝ωl2·xl(k)＝ωl21xX1+ωl22xX2+ωl23xY1+ωl24xY2
wherein ,ωl1 and ωl2Is the l gain vector.
Furthermore, in one implementation of this embodiment, the non-linear state vector F of the l-axisl[xl(k)]It may also be one of the following formulas:
Figure BDA0002433244320000061
Figure BDA0002433244320000062
wherein ,hlIs the width of the Gaussian kernel function, clIs the gaussian kernel function center point.
Further, in an implementation manner of this embodiment, the neural network model is built based on the planar motor nonlinear dynamic model, and the model parameters of the neural network model are model parameters of the planar motor nonlinear dynamic model. The input item of the neural network model is state information of the planar motor, and the output item of the neural network model is position information of the planar motor at the next moment. It can be understood that the neural network model is established based on a state equation of the nonlinear dynamic model of the planar motor, and the input items are control quantity and state information in the state equation; the output item is a state equation to determine position information in the expected state information, and a model function corresponding to the neural network model is a nonlinear state vector F of an I axis in the nonlinear dynamic model of the planar motorl[xl(k)]The model parameter of the neural network model is a coefficient matrix G of an l axisl and Hl. Thus, the neural network model is trained as model parameters G for the neural network modell and HlDetermining a coefficient matrix G of the nonlinear dynamic model of the planar motorl and HlThe model accuracy of the nonlinear dynamic model of the planar motor is improved, and therefore the accuracy of the nonlinear dynamic model of the planar motor for controlling the planar motor is improved.
And S20, training the neural network model based on a preset training sample set.
Specifically, the preset training sample set includes a plurality of training samples, each training sample includes training data and real state information corresponding to the training data, and the training data includes state information and a control quantity. The real state information is state information of the next moment of the training data, and the state information comprises speed information and position information. Therefore, the training sample data comprises speed information at the current moment, position information at the current moment and control quantity information at the current moment, and the output item of the neural network model is predicted state information at the next moment.
In an implementation manner of this embodiment, the obtaining process of the preset training sample set may be: five samples are collected, the corresponding track and speed of each sample are different, and each sample at least comprises 1 ten thousand sets of sequence data, for example, taking the step length of 1ms as an example, data with the time length of at least 10s needs to be collected. In addition, one group of five samples is used for training the neural network-based planar motor nonlinear dynamic model, and the remaining four groups of samples are used for testing the neural network-based planar motor nonlinear dynamic model. In addition, each set of training data comprises an l-axis control quantity, an l-axis position and an l-axis speed, wherein the l-axis speed is calculated based on the l-axis position by using a numerical differentiation algorithm, and the l-axis control quantity is the control quantity output by the controller in the planar motor control system, so that the nonlinear dynamic model of the planar motor reflects the constraint characteristic of actuator saturation through the control quantity output by the controller in the planar motor control system.
In an implementation manner of this embodiment, as shown in fig. 3, the training the neural network model based on the preset training sample set specifically includes:
s21, inputting the training data into the neural network model, and extracting the prediction state information of the training data through the neural network model;
s22, determining a loss function corresponding to the training data based on the predicted state information and the real state information corresponding to the training data;
s23, training the neural network model based on the loss function.
Specifically, the predicted state information is the expected state information at the next time, for example, if the state information in the training data is the state information at time k, then the predicted state information is the expected state information at time k + 1. It can be understood that, when the state information in the training data is k, the actual state information is seen, the actual state information is the actual state information at the time k +1, and the predicted state information is the expected state information at the time k + 1. Therefore, the loss function corresponding to the training data can be determined based on the predicted state information and the real state information corresponding to the training data, and the neural network model can be trained by adopting the loss function.
In an alternative embodiment, the expression of the loss function may be:
Figure BDA0002433244320000071
wherein J is a loss function, xreal_l(k) For true state information, xl(k) To predict the state information, k denotes the kth time.
Further, after the loss function is determined, a gradient descent method can be adopted for learning to correct the network parameters of the neural network model to obtain model parameters meeting preset conditions, and further obtain nonlinear model parameters G of the planar motorl and Hl. The model parameter meets the preset condition that the loss function value is smaller than a preset threshold value or the training times of the neural network model reach a preset time threshold value. In addition, the non-linear state vector F due to the l-axisl[xl(k)]The coupling function in (1) contains a coupling coefficient, so that when the network model is corrected, the coupling coefficient is corrected to improve the nonlinear state vector F of the l axisl[xl(k)]The accuracy of the nonlinear dynamic model of the planar motor is further improved.
The G isl、HlAnd a coupling coefficient omegalThe adjustment formula of (c) may be:
Gl(k+1)＝Gl(k)-ηEl(k)Fl[x(k)]+α[Gl(k)-Gl(k-1)]
Hl(k+1)＝Hl(k)-ηEl(k)ul(k)+α[Hl(k)-Hl(k-1)]
wl11(k+1)＝wl11(k)-η{g1(k)x1(k)+g3(k)x1(k)}
wl12(k+1)＝wl12(k)-η{g1(k)x2(k)+g3(k)x2(k)}
wl13(k+1)＝wl13(k)-η{g1(k)x3(k)+g3(k)x3(k)}
wl14(k+1)＝wl14(k)-η{g1(k)x4(k)+g3(k)x4(k)}
wl21(k+1)＝wl21(k)-η{g2(k)x1(k)+g4(k)x1(k)}
wl22(k+1)＝wl22(k)-η{g2(k)x2(k)+g4(k)x2(k)}
wl23(k+1)＝wl23(k)-η{g2(k)x3(k)+g4(k)x3(k)}
wl24(k+1)＝wl24(k)-η{g2(k)x4(k)+g4(k)x4(k)}
wherein η is learning rate, α is momentum factor, g1、g2、g3、g4Are respectively as
g1(k)＝El1(k)Gl11Al1al1[1-fl1(xl(k))2]
g2(k)＝El1(k)Gl12Al2al2[1-fl2(xl(k))2]
g3(k)＝El2(k)Gl21Al1al1[1-fl1(xl(k))2]
g4(k)＝El2(k)Gl22Al2al2[1-fl2(xl(k))2]
Further, in this embodiment, the neural network model includes a mapping unit and a linear transformation unit; the inputting the training data into the neural network model, and the extracting the predicted state information of the training data through the neural network model specifically includes:
inputting the state information into the mapping unit, and outputting a coupling parameter through the mapping unit;
and inputting the coupling parameters and the control quantity into the linear transformation unit, and outputting the prediction state information of the training data through the linear transformation unit.
Specifically, the linear transformation unit may include at least one hidden layer, and the number of nodes of the last hidden layer is greater than or equal to 2, and the coupling parameter and the control quantity are transmitted through the convolutional layer, and the nonlinear state vector F of the l-axis is transmitted through the convolutional layerl[xl(k)]Predicted speed information and position information are obtained. For example, as shown in FIG. 4, the linear transformation unit includes a convolution layer, and F is set when the number of nodes of the convolution layer is 2l[xl(k)]∈R2×1、Gl∈R2×2(ii) a When the number of hidden layer nodes is n, Fl[xl(k)]∈Rn×1、Gl∈R2×n(ii) a As shown in FIG. 5, the hidden layer can be extended to a neural network with m layers and n nodes in each layer. The specific values of m and n are determined according to modeling requirements.
And S30, taking the model parameters of the trained neural network model as the model parameters of the nonlinear dynamic model of the planar motor to obtain the nonlinear dynamic model of the planar motor.
Specifically, after the model parameters of the neural network model meet preset conditions, G in the neural network model is obtainedl and HlAnd G isl and HlThe model parameter is used as the model parameter of the nonlinear dynamic model of the planar motor, thereby improving the precision and the control performance of the nonlinear dynamic model of the planar motor and being suitable for the design of the controller of the planar motor. In addition, the coupling characteristic of two-dimensional motion is considered in the nonlinear dynamic model of the planar motor, the nonlinear dynamic model with independent X axis and Y axis is established, and the precision of the planar motor model is improvedThe controller structure of the planar motor with the coupling characteristic is simplified due to the degree and the control performance. Meanwhile, the nonlinear dynamic model of the planar motor adopts the control quantity output by the controller as an input item of the neural network model, so that the model parameters meet the constraint characteristic of the saturation of the nonlinear actuator, the precision and the control performance of the planar motor model are improved, the structure of the controller considering the saturation of the actuator is simplified, and the nonlinear dynamic model of the planar motor is more suitable for an actual planar motor system.
Based on the above construction method of the nonlinear model of the planar motor, this embodiment provides a control method of the planar motor, and the nonlinear model of the planar motor constructed by applying the construction method of the nonlinear model of the planar motor described in the above embodiment includes:
determining expected position information of the planar motor at the next moment based on the nonlinear model of the planar motor;
determining a control amount of the planar motor based on the expected position information and actual position information of the planar motor;
and applying the control quantity to the planar motor to control the planar motor.
Based on the above method for constructing a nonlinear model of a planar motor, the present embodiment provides a computer-readable storage medium, which stores one or more programs that can be executed by one or more processors to implement the steps in the method for constructing a nonlinear model of a planar motor according to the above embodiment.
Based on the above construction method of the nonlinear model of the planar motor, the present invention further provides a terminal device, as shown in fig. 6, which includes at least one processor (processor) 20; a display screen 21; and a memory (memory)22, and may further include a communication Interface (Communications Interface)23 and a bus 24. The processor 20, the display 21, the memory 22 and the communication interface 23 can communicate with each other through the bus 24. The display screen 21 is configured to display a user guidance interface preset in the initial setting mode. The communication interface 23 may transmit information. The processor 20 may call logic instructions in the memory 22 to perform the methods in the embodiments described above.
Furthermore, the logic instructions in the memory 22 may be implemented in software functional units and stored in a computer readable storage medium when sold or used as a stand-alone product.
The memory 22, which is a computer-readable storage medium, may be configured to store a software program, a computer-executable program, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure. The processor 20 executes the functional application and data processing, i.e. implements the method in the above-described embodiments, by executing the software program, instructions or modules stored in the memory 22.
The memory 22 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal device, and the like. Further, the memory 22 may include a high speed random access memory and may also include a non-volatile memory. For example, a variety of media that can store program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk, may also be transient storage media.
In addition, the specific processes loaded and executed by the storage medium and the instruction processors in the terminal device are described in detail in the method, and are not stated herein.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. A method for constructing a nonlinear model of a planar motor is characterized by comprising the following steps:
establishing a nonlinear dynamic model of a planar motor, and establishing a neural network model based on the nonlinear dynamic model of the planar motor, wherein model parameters of the neural network model are model parameters of the nonlinear dynamic model of the planar motor;
training the neural network model based on a preset training sample set;
and taking the model parameters of the trained neural network model as the model parameters of the nonlinear dynamic model of the planar motor to obtain the nonlinear dynamic model of the planar motor.
2. The method for constructing the nonlinear model of the planar motor according to claim 1, wherein the preset training sample set includes a plurality of training samples, each training sample includes training data and real state information corresponding to the training data, and the training data includes state information and a control quantity.
3. The method for constructing the nonlinear model of the planar motor according to claim 2, wherein the real state information is state information of a next moment of the training data, wherein the state information includes speed information and position information.
4. The method for constructing the nonlinear model of the planar motor according to claim 2 or 3, wherein the training of the neural network model based on the preset training sample set specifically includes:
inputting the training data into the neural network model, and extracting the prediction state information of the training data through the neural network model;
determining a loss function corresponding to the training data based on the predicted state information and the real state information corresponding to the training data;
training the neural network model based on the loss function.
5. The construction method of the nonlinear model of the planar motor according to claim 4, wherein the neural network model comprises a mapping unit and a linear transformation unit; inputting the training data into the neural network model, and extracting the predicted state information of the training data through the neural network model specifically comprises:
inputting the state information into the mapping unit, and outputting a coupling parameter through the mapping unit;
and inputting the coupling parameters and the control quantity into the linear transformation unit, and outputting the prediction state information of the training data through the linear transformation unit.
6. The method for constructing the nonlinear model of the planar motor according to claim 4, wherein the loss function is:
Figure FDA0002433244310000021
wherein J is a loss function, xreal_l(k) For true state information, xl(k) To predict the state information, k denotes the kth time.
7. The method for constructing the nonlinear dynamic model of the planar motor according to claim 1, wherein the establishing the nonlinear dynamic model of the planar motor comprises the following steps:
Figure FDA0002433244310000022
wherein ,
Figure FDA0002433244310000023
xl1(k) position information at time k on the l-axis, xl2(k) Speed information at time k on the l-axis, yl(k) Position information output for the model at time k, Cl＝[10]，Gl and HlIs a coefficient matrix of the l-axis, Fl[xl(k)]A non-linear state vector of the l-axis, ul(k) In order to control the amount of the liquid,
Figure FDA0002433244310000024
xl1(k +1) is position information of the time k +1 on the l-axis, xl2(k +1) is velocity information at the time of the l-axis k + 1.
8. A method for controlling a planar motor, wherein the planar motor nonlinear model is constructed by applying the method for constructing a planar motor nonlinear model according to any one of claims 1 to 7, the method comprising:
determining expected position information of the planar motor at the next moment based on the nonlinear model of the planar motor;
determining a control amount of the planar motor based on the expected position information and actual position information of the planar motor;
and applying the control quantity to the planar motor to control the planar motor.
9. A computer-readable storage medium storing one or more programs, which are executable by one or more processors, to implement the steps in the method for constructing a nonlinear model of a planar motor according to any one of claims 1 to 7.
10. A terminal device, comprising: a processor, a memory, and a communication bus; the memory has stored thereon a computer readable program executable by the processor;
the communication bus realizes connection communication between the processor and the memory;
the processor, when executing the computer readable program, implements the steps in the method of constructing a nonlinear model of a planar motor according to any one of claims 1 to 7.
CN202010243219.XA 2020-03-31 2020-03-31 Construction method of nonlinear model of planar motor Active CN111460732B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010243219.XA CN111460732B (en) 2020-03-31 2020-03-31 Construction method of nonlinear model of planar motor

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010243219.XA CN111460732B (en) 2020-03-31 2020-03-31 Construction method of nonlinear model of planar motor

Publications (2)

Publication Number Publication Date
CN111460732A true CN111460732A (en) 2020-07-28
CN111460732B CN111460732B (en) 2023-05-16

Family

ID=71683434

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010243219.XA Active CN111460732B (en) 2020-03-31 2020-03-31 Construction method of nonlinear model of planar motor

Country Status (1)

Country Link
CN (1) CN111460732B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019062413A1 (en) * 2017-09-30 2019-04-04 Oppo广东移动通信有限公司 Method and apparatus for managing and controlling application program, storage medium, and electronic device
CN110162799A (en) * 2018-11-28 2019-08-23 腾讯科技（深圳）有限公司 Model training method, machine translation method and relevant apparatus and equipment
CN110829934A (en) * 2019-11-27 2020-02-21 华南理工大学 Permanent magnet alternating current servo intelligent control system based on definite learning and mode control

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019062413A1 (en) * 2017-09-30 2019-04-04 Oppo广东移动通信有限公司 Method and apparatus for managing and controlling application program, storage medium, and electronic device
CN110162799A (en) * 2018-11-28 2019-08-23 腾讯科技（深圳）有限公司 Model training method, machine translation method and relevant apparatus and equipment
CN110829934A (en) * 2019-11-27 2020-02-21 华南理工大学 Permanent magnet alternating current servo intelligent control system based on definite learning and mode control

Also Published As

Publication number Publication date
CN111460732B (en) 2023-05-16

Similar Documents

Publication Publication Date Title
CN107169563B (en) Processing system and method applied to two-value weight convolutional network
CN112766596B (en) Construction method of building energy consumption prediction model, energy consumption prediction method and device
CN112000893B (en) Resident region prediction method, device, equipment and storage medium
CN106850289B (en) Service combination method combining Gaussian process and reinforcement learning
CN109514553A (en) A kind of method, system and the equipment of the mobile control of robot
CN110991630A (en) Convolutional neural network processor for edge calculation
CN112860869B (en) Dialogue method, device and storage medium based on hierarchical reinforcement learning network
CN111640296B (en) Traffic flow prediction method, system, storage medium and terminal
CN112580733A (en) Method, device and equipment for training classification model and storage medium
CN115293623A (en) Training method and device for production scheduling model, electronic equipment and medium
CN111639523B (en) Target detection method, device, computer equipment and storage medium
CN110782023B (en) Reduction residual module porous convolution architecture network and rapid semantic segmentation method
CN117351299A (en) Image generation and model training method, device, equipment and storage medium
CN111460732B (en) Construction method of nonlinear model of planar motor
CN116048708B (en) Software window adjusting method, system, equipment and medium based on artificial intelligence
CN110135561B (en) Real-time online aircraft AI neural network system
WO2022127603A1 (en) Model processing method and related device
CN116306808A (en) Convolutional neural network compression method and device combining dynamic pruning and conditional convolution
CN115952834A (en) Batch layer normalization method and device applied to Transformer
CN115545168A (en) Dynamic QoS prediction method and system based on attention mechanism and recurrent neural network
Sun et al. Super-resolution imaging using convolutional neural networks
CN117540677A (en) Method and device for acquiring power amplifier model and power amplifier model
CN110609971B (en) Calibration multiple regression network construction method
CN113139490A (en) Image feature matching method and device, computer equipment and storage medium
CN116090573B (en) Method, device, equipment and storage medium for simulating non-local quantum operation

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant